On Sun, May 24, 2015 at 3:51 PM, Ish Sookun <ish_at_hacklog.in> wrote:
> Hi Logan,
>
> On Sun, May 24, 2015 at 7:41 PM, Loganaden Velvindron <loganaden_at_gmail.com>
> wrote:
>>
>> On Sun, May 24, 2015 at 3:31 PM, S Moonesamy <sm+mu_at_elandsys.com> wrote:
>>
>> We have a large latency and increasing bandwidth. At the Linux kernel
>> level, you can see work sponsored by Google to make networking
>> "faster". It's very important , as it gives the user a better
>> experience.
>
>
> Can you give a proper reference?
>
https://tools.ietf.org/html/rfc6928
>>
>>
>> I strongly believe that Mauritius and Africa could have benefitted by
>> participating by testing and giving feedback on how the changes made
>> by US/EU developers to the Linux networking stack behave in our
>> context. As a bonus, it would be nice to be able to tweak the changes
>> for Mauritius ;)
>
>
> I didn't grasp this part. Does the code behave differently in our
> region/context?
Yes, it does, due to the conditions of our network links here.
See page 5.
http://www.ietf.org/proceedings/78/slides/iccrg-3.pdf
>
> If this is true, it interests me & I would like to participate. What do you
> advise?
>
There was a long and interesting discussion regarding the pros and
cons of increasing the initial TCP window, when I proposed to
implement it in NetBSD :) It gave me different perspectives on how
people see the problem.
http://comments.gmane.org/gmane.os.netbsd.devel.network/13518
You can set up a NetBSD box, and test the patch :)
Received on Sun May 24 2015 - 19:27:12 PST